![SonomaEye-full-color-vector[2].png](https://static.wixstatic.com/media/bed00b_da20b9cc704044b0b048ac8ef9190470~mv2_d_3024_2195_s_2.png/v1/fill/w_177,h_123,al_c,q_85,usm_0.66_1.00_0.01,enc_avif,quality_auto/SonomaEye-full-color-vector%5B2%5D.png)
Thank you for your submission!
Diagnostic journey, care coordination and patient subtypes for Sjögren’s Disease: A complicated path to disease diagnosis
Alexander Keenan, MA, MHP
Presenter:
Alexander Keenan(1), Esen Akpek(2), Elizabeth Adamson(1), Federico Zazzetti(1), Soumya D Chakravarty(1,3), Sujung Choi(1), Chuxuan Yang(1) Yafen Huang(1), Nora Sandorfi(4)
Authors:
Affiliation:
1. Johnson & Johnson, Horsham, PA, USA
2. Foster Center for Ocular Immunology, Duke University School of Medicine, Durham, NC, USA
3. Drexel University School of Medicine, Philadelphia, PA, USA
4. University of Pennsylvania, Philadelphia, PA, USA
Purpose
Understanding patient characteristics, comorbidities, healthcare utilization, and treatment journey through different healthcare specialists allows for a greater understanding regarding the challenges associated with any disease process and provide opportunities to optimize patient care. Our aim here is to investigate the patient journey prior to Sjögren’s Disease (SjD) diagnosis, characterize unique patient subtypes based on clinical presentation and diagnostic journey for both primary and associated SjD.
Methods
A retrospective observational cohort study, using patient level sequence data for clustering SjD patients into subtypes based on similarities in their diagnostic journeys and disease manifestations for newly diagnosed SjD patients from IQVIA LAAD Extended Immunology dataset spanning 3Q 2018 to 3Q 2023. Patient subtypes were characterized according to demographics, clinical features, healthcare utilization, care coordination, and other factors based on the Leiden community detection method. This allows for identification of distinct patient subtypes based on clinical presentation and diagnostic journeys. The algorithm grouped patients into clusters, revealing patterns in symptom progression and healthcare provider interactions.
Results
The cluster analysis identified 5,126 newly diagnosed Sjogren’s patients (88% female and 12% male) classified into four dominant subtypes: muscle/joint (36%), mood (25%), CNS prevalent (18%), and autoimmune/eye (21%). The average age at diagnosis was 62.67 (SD: 13.67), with an average time to diagnosis of 1520 days (SD 17`) or 4.16 years. The average patient had 65.7 unique HCP visits and 9.9 HCP specialists interactions from first symptom to confirmed Sjögren’s diagnosis. Rheumatologists diagnosed 36% of patients, followed by eye specialists (21%), general practitioners (19%), and other HCPs (24%).
Conclusion
This novel machine learning approach was able to capture and highlight the complexity of Sjögren's Disease journey and the various disease manifestations. Through cluster analysis, 4 different SjD patient subtypes were identified based on symptom manifestation prior to diagnosis, allowing for a greater understanding of the complicated diagnostic journey. The better understanding of the current referral patterns provide an opportunity for education of health care practitioners regarding the burden of Sjögren's Disease, and the opportunity to narrow the time from symptom onset to diagnosis.